@stll/folio-core 0.60.0 → 0.62.0

package/dist/docx/runParser.js

@@ -1,4 +1,5 @@
 import { parseHorizontalScalePercent } from "../utils/horizontalScale.js";
+import { sliceText } from "../utils/sliceText.js";
 import { NO_MODELLED_ATTRIBUTES, attributeRemainder } from "./attributeRemainder.js";
 import { CAPTURE, dispatchChildrenWithContext, keptUnless, ownedElsewhere, sequencePositions } from "./containerChildren.js";
 import { parseDiagramPreview } from "./diagramPreview.js";
@@ -10,6 +11,7 @@ import { EmphasisMarkSchema, FontHintSchema, FontThemeSchema, HighlightColorSche
 import { preserveRunChild } from "./preservedRunContent.js";
 import { standalonePreviewLedger } from "./previewBudget.js";
 import { parseBoundaryJoins } from "./reviewResolutionProvenance.js";
+import { isEmptyRunPropertiesAttribute, omitsPreviousRunProperties, retainsEmptyRunProperties } from "./runPropertyPresence.js";
 import { parseShading } from "./shadingParser.js";
 import { captureShapeAlternateContent } from "./shapeAlternateContent.js";
 import { parseShapeFromDrawing, shouldPreserveRawShapeDrawing } from "./shapeParser.js";
@@ -48,7 +50,11 @@ import { DRAWING_RAW_XML_MODES } from "@stll/docx-core/model";
 * attacker-sized string per run.
 */
 const MAX_LANGUAGE_TAG_LENGTH = 35;
-const truncateLanguageTag = (value) => value === void 0 ? void 0 : value.slice(0, MAX_LANGUAGE_TAG_LENGTH);
+const truncateLanguageTag = (value) => value === void 0 ? void 0 : sliceText({
+	text: value,
+	units: "codeUnits",
+	end: MAX_LANGUAGE_TAG_LENGTH
+});
 function parseColorValue({ rgb, themeColor, themeTint, themeShade, element }) {
 	const color = {};
 	if (rgb && rgb !== "auto") color.rgb = rgb;
@@ -429,7 +435,7 @@ function parseRunProperties(rPr, theme, owner) {
 	});
 	if (preserved) formatting.preserved = preserved;
 	const children = getChildElements(rPr);
-	if (owner === RUN_PROPERTY_OWNERS.paragraphMark && Object.keys(formatting).length === 0 && children.length > 0 && children.every((child) => WORDPROCESSINGML_NAMESPACE_URIS.has(getNamespaceUri(child) ?? "") && Object.hasOwn(OWNED_BY_A_SIBLING_RECORD.paragraphMark, getLocalName(child.name)))) return;
+	if (owner !== RUN_PROPERTY_OWNERS.standalone && Object.keys(formatting).length === 0 && children.length > 0 && children.every((child) => WORDPROCESSINGML_NAMESPACE_URIS.has(getNamespaceUri(child) ?? "") && Object.hasOwn(OWNED_BY_A_SIBLING_RECORD[owner], getLocalName(child.name)))) return;
 	return formatting;
 }
 function parseRunPropertyChanges(rPr, theme, currentFormatting) {
@@ -443,7 +449,7 @@ function parseRunPropertyChanges(rPr, theme, currentFormatting) {
 		const boundaryJoins = parseBoundaryJoins(changeElement);
 		if (boundaryJoins !== void 0) change.boundaryJoins = boundaryJoins;
 		const previousFormatting = parseRunProperties(previousRPr, theme, RUN_PROPERTY_OWNERS.standalone);
-		if (previousFormatting) change.previousFormatting = previousFormatting;
+		if (previousFormatting && (!omitsPreviousRunProperties(changeElement) || Object.keys(previousFormatting).length > 0)) change.previousFormatting = previousFormatting;
 		if (currentFormatting) change.currentFormatting = currentFormatting;
 		return change;
 	});
@@ -789,7 +795,7 @@ function parseRun(node, _styles, theme, rels = null, media = null, rootXmlns = {
 	};
 	const rPr = findChild(node, "w", "rPr");
 	if (rPr) {
-		const formattingResult = parseRunProperties(rPr, theme, RUN_PROPERTY_OWNERS.run);
+		const formattingResult = parseRunProperties(rPr, theme, RUN_PROPERTY_OWNERS.run) ?? (retainsEmptyRunProperties(node) ? {} : void 0);
 		if (formattingResult) run.formatting = formattingResult;
 		const propertyChangesResult = parseRunPropertyChanges(rPr, theme, run.formatting);
 		if (propertyChangesResult) run.propertyChanges = propertyChangesResult;
@@ -798,8 +804,8 @@ function parseRun(node, _styles, theme, rels = null, media = null, rootXmlns = {
 	const remainder = attributeRemainder({
 		element: node,
 		modelled: NO_MODELLED_ATTRIBUTES
-	});
-	if (remainder) run.preservedAttributes = remainder;
+	})?.filter((attribute) => !isEmptyRunPropertiesAttribute(attribute));
+	if (remainder?.length) run.preservedAttributes = remainder;
 	return run;
 }
 /**

package/dist/docx/runPropertyPresence.d.ts

@@ -0,0 +1,11 @@
+import { document_d_exports } from "../types/document.js";
+import { XmlElement } from "./xmlParser.js";
+//#region src/docx/runPropertyPresence.d.ts
+declare const EMPTY_RUN_PROPERTIES_ATTRIBUTE = "emptyRunProperties";
+declare const retainsEmptyRunProperties: (element: XmlElement) => boolean;
+declare const omitsPreviousRunProperties: (element: XmlElement) => boolean;
+declare const serializePreviousRunProperties: (properties: document_d_exports.TextFormatting | undefined) => string;
+declare const serializeEmptyRunProperties: (run: document_d_exports.Run) => string[];
+declare const isEmptyRunPropertiesAttribute: ({ namespace, name }: Pick<document_d_exports.PreservedAttribute, "namespace" | "name">) => boolean;
+//#endregion
+export { EMPTY_RUN_PROPERTIES_ATTRIBUTE, isEmptyRunPropertiesAttribute, omitsPreviousRunProperties, retainsEmptyRunProperties, serializeEmptyRunProperties, serializePreviousRunProperties };

package/dist/docx/runPropertyPresence.js

@@ -0,0 +1,14 @@
+import { OOXML_NAMESPACES } from "./serializer/partNamespaces.js";
+import { getAttributeByNamespaceUri } from "./xmlParser.js";
+import { EMPTY_PROPERTY_SETS } from "@stll/docx-core/ops";
+//#region src/docx/runPropertyPresence.ts
+const EMPTY_RUN_PROPERTIES_ATTRIBUTE = "emptyRunProperties";
+const PREVIOUS_PROPERTIES_ATTRIBUTE = "previousProperties";
+const HISTORY_NAMESPACES = /* @__PURE__ */ new Set([OOXML_NAMESPACES.folio.uri]);
+const retainsEmptyRunProperties = (element) => getAttributeByNamespaceUri(element, HISTORY_NAMESPACES, EMPTY_RUN_PROPERTIES_ATTRIBUTE) === EMPTY_PROPERTY_SETS.KEEP;
+const omitsPreviousRunProperties = (element) => getAttributeByNamespaceUri(element, HISTORY_NAMESPACES, PREVIOUS_PROPERTIES_ATTRIBUTE) === EMPTY_PROPERTY_SETS.OMIT;
+const serializePreviousRunProperties = (properties) => properties === void 0 ? ` folio:${PREVIOUS_PROPERTIES_ATTRIBUTE}="${EMPTY_PROPERTY_SETS.OMIT}"` : "";
+const serializeEmptyRunProperties = (run) => run.propertyChanges?.length && run.formatting !== void 0 && Object.keys(run.formatting).length === 0 ? [`folio:${EMPTY_RUN_PROPERTIES_ATTRIBUTE}="${EMPTY_PROPERTY_SETS.KEEP}"`] : [];
+const isEmptyRunPropertiesAttribute = ({ namespace, name }) => namespace === OOXML_NAMESPACES.folio.uri && name === "emptyRunProperties";
+//#endregion
+export { EMPTY_RUN_PROPERTIES_ATTRIBUTE, isEmptyRunPropertiesAttribute, omitsPreviousRunProperties, retainsEmptyRunProperties, serializeEmptyRunProperties, serializePreviousRunProperties };

package/dist/docx/selectiveSave.js

@@ -1,3 +1,4 @@
+import { decodeText, encodeUtf8 } from "../utils/textEncoding.js";
 import { parseCommentsExtended } from "./commentParser.js";
 import { withoutOrphanCommentRanges } from "./commentRangeIntegrity.js";
 import { hasUnsynthesizedReplyRanges } from "./commentReplyMarkers.js";
@@ -46,7 +47,7 @@ async function patchNoteParts(zip, doc, candidateIds, updates) {
 	const patchPart = async (conventionalLowerPath, serialize) => {
 		const file = findPart(conventionalLowerPath);
 		if (!file) return true;
-		const originalXml = await file.async("text");
+		const originalXml = await file.async("uint8array").then((bytes) => decodeText(bytes, { ignoreBOM: true }));
 		const originalIds = collectParaIds(originalXml);
 		const partIds = /* @__PURE__ */ new Set();
 		for (const id of remainingIds) if (originalIds.has(id)) partIds.add(id);
@@ -93,7 +94,7 @@ async function patchCommentsExtended(zip, plan, updates) {
 	const desiredXml = serializeCommentsExtended(plan);
 	const existing = findZipEntryCaseInsensitive(zip, COMMENTS_EXTENDED_PART_LOWER);
 	if (!desiredXml) return existing === null;
-	const baselineInfo = existing ? parseCommentsExtended(await existing.async("text")) : /* @__PURE__ */ new Map();
+	const baselineInfo = existing ? parseCommentsExtended(await existing.async("uint8array").then((bytes) => decodeText(bytes, { ignoreBOM: true }))) : /* @__PURE__ */ new Map();
 	if (commentsExtendedInfoEqual(parseCommentsExtended(desiredXml), baselineInfo)) return true;
 	updates.set(existing?.name ?? "word/commentsExtended.xml", desiredXml);
 	return ensureCommentsExtendedPackaging(zip, updates);
@@ -106,12 +107,12 @@ async function patchCommentsExtended(zip, plan, updates) {
 */
 async function ensureCommentsExtendedPackaging(zip, updates) {
 	const ctPath = "[Content_Types].xml";
-	const ctXml = updates.get(ctPath) ?? await zip.file(ctPath)?.async("text");
+	const ctXml = updates.get(ctPath) ?? await zip.file(ctPath)?.async("uint8array").then((bytes) => decodeText(bytes, { ignoreBOM: true }));
 	if (ctXml === void 0) return false;
 	const nextCt = addCommentsExtendedOverride(ctXml);
 	if (nextCt !== ctXml) updates.set(ctPath, nextCt);
 	const relsPath = "word/_rels/document.xml.rels";
-	const relsXml = updates.get(relsPath) ?? await zip.file(relsPath)?.async("text");
+	const relsXml = updates.get(relsPath) ?? await zip.file(relsPath)?.async("uint8array").then((bytes) => decodeText(bytes, { ignoreBOM: true }));
 	if (relsXml === void 0) return false;
 	const nextRels = addCommentsExtendedRelationship(relsXml);
 	if (nextRels !== relsXml) updates.set(relsPath, nextRels);
@@ -146,7 +147,7 @@ async function patchNumberingPart(zip, doc, updates) {
 		}
 	}
 	if (!file) return false;
-	const originalXml = await file.async("text");
+	const originalXml = await file.async("uint8array").then((bytes) => decodeText(bytes, { ignoreBOM: true }));
 	const baselineXml = serializeNumberingXml(parseNumbering(originalXml).definitions);
 	const currentXml = serializeNumberingXml(numbering);
 	if (currentXml === baselineXml) return true;
@@ -177,7 +178,7 @@ const queueSettingsUpdates = async (zip, updates) => {
 	const settingsFile = zip.file("word/settings.xml");
 	if (!settingsFile) return;
 	const relsFile = zip.file("word/_rels/settings.xml.rels");
-	const filtered = withoutAttachedTemplate(await settingsFile.async("text"), await relsFile?.async("text"));
+	const filtered = withoutAttachedTemplate(await settingsFile.async("uint8array").then((bytes) => decodeText(bytes, { ignoreBOM: true })), await relsFile?.async("uint8array").then((bytes) => decodeText(bytes, { ignoreBOM: true })));
 	if (filtered.settingsXml !== void 0) updates.set("word/settings.xml", filtered.settingsXml);
 	if (filtered.relsXml !== void 0) updates.set("word/_rels/settings.xml.rels", filtered.relsXml);
 };
@@ -212,12 +213,12 @@ async function attemptSelectiveSave(doc, originalBuffer, options) {
 			for (const [path, file] of Object.entries(zip.files)) {
 				const lowerPath = path.toLowerCase();
 				if (file.dir || !lowerPath.startsWith("word/") || !lowerPath.endsWith(".xml")) continue;
-				sourceParts.set(path, await file.async("text"));
+				sourceParts.set(path, await file.async("uint8array").then((bytes) => decodeText(bytes, { ignoreBOM: true })));
 			}
 			for (const [path, xml] of normalizeImportedNumericIds(sourceParts)) if (xml !== sourceParts.get(path)) writeZipPart({
 				zip,
 				path,
-				data: xml
+				data: encodeUtf8(xml)
 			});
 		}
 		if (options.bodyAuthority === "canonical") await publishCanonicalImageResources({
@@ -229,7 +230,7 @@ async function attemptSelectiveSave(doc, originalBuffer, options) {
 		if (changedParaIds.size > 0 || structuralChange) {
 			const docXmlFile = zip.file("word/document.xml");
 			if (!docXmlFile) return null;
-			const originalDocXml = await docXmlFile.async("text");
+			const originalDocXml = await docXmlFile.async("uint8array").then((bytes) => decodeText(bytes, { ignoreBOM: true }));
 			const serializedDocXml = serializeDocument(doc, readRootNamespaceBindings(originalDocXml), {
 				xml: originalDocXml,
 				bodyAuthority: options.bodyAuthority,
@@ -239,7 +240,7 @@ async function attemptSelectiveSave(doc, originalBuffer, options) {
 			const originalBodyParaIds = structuralChange ? collectParaIds(originalDocXml) : void 0;
 			if (structuralChange && doc.package.numbering) {
 				const numberingFile = findZipEntryCaseInsensitive(zip, "word/numbering.xml");
-				const added = collectAddedNumberingDefs(numberingFile ? serializeNumberingXml(parseNumbering(await numberingFile.async("text")).definitions) : "", serializeNumberingXml(doc.package.numbering));
+				const added = collectAddedNumberingDefs(numberingFile ? serializeNumberingXml(parseNumbering(await numberingFile.async("uint8array").then((bytes) => decodeText(bytes, { ignoreBOM: true }))).definitions) : "", serializeNumberingXml(doc.package.numbering));
 				if (added.abstractNums.size > 0 || added.nums.size > 0) return null;
 			}
 			const bodyChangedIds = /* @__PURE__ */ new Set();
@@ -262,7 +263,7 @@ async function attemptSelectiveSave(doc, originalBuffer, options) {
 		}
 		const sourceCommentsFile = findZipEntryCaseInsensitive(zip, "word/comments.xml");
 		const commentsPartPath = sourceCommentsFile?.name ?? "word/comments.xml";
-		const sourceCommentsXml = sourceCommentsFile ? await sourceCommentsFile.async("text") : void 0;
+		const sourceCommentsXml = sourceCommentsFile ? await sourceCommentsFile.async("uint8array").then((bytes) => decodeText(bytes, { ignoreBOM: true })) : void 0;
 		const commentsUnchanged = sourceCommentsXml !== void 0 && canReplayCommentsPart(doc, sourceCommentsXml);
 		if (!commentsUnchanged && (hasComments || ownsCommentsPart || sourceCommentsFile)) {
 			const commentsXml = serializeComments(commentPlan, sourceCommentsXml === void 0 ? void 0 : readRootNamespaceBindings(sourceCommentsXml));
@@ -271,13 +272,13 @@ async function attemptSelectiveSave(doc, originalBuffer, options) {
 		if (hasComments || ownsCommentsPart) {
 			const ctFile = zip.file("[Content_Types].xml");
 			if (ctFile) {
-				const ctXml = await ctFile.async("text");
+				const ctXml = await ctFile.async("uint8array").then((bytes) => decodeText(bytes, { ignoreBOM: true }));
 				if (!ctXml.toLowerCase().includes("/word/comments.xml")) updates.set("[Content_Types].xml", ctXml.replace("</Types>", `<Override PartName="/${commentsPartPath}" ContentType="${COMMENTS_CONTENT_TYPE}"/></Types>`));
 			}
 			const relsPath = "word/_rels/document.xml.rels";
 			const relsFile = zip.file(relsPath);
 			if (relsFile) {
-				const relsXml = await relsFile.async("text");
+				const relsXml = await relsFile.async("uint8array").then((bytes) => decodeText(bytes, { ignoreBOM: true }));
 				if (!relsXml.toLowerCase().includes("comments.xml")) {
 					const maxId = findMaxRId(relsXml);
 					updates.set(relsPath, relsXml.replace("</Relationships>", `<Relationship Id="rId${maxId + 1}" Type="${RELATIONSHIP_TYPES.comments}" Target="${commentsPartPath.slice(5)}"/></Relationships>`));
@@ -307,7 +308,7 @@ async function attemptSelectiveSave(doc, originalBuffer, options) {
 		await queueSettingsUpdates(zip, updates);
 		const corePropsFile = zip.file("docProps/core.xml");
 		if (corePropsFile) {
-			const corePropsXml = await corePropsFile.async("text");
+			const corePropsXml = await corePropsFile.async("uint8array").then((bytes) => decodeText(bytes, { ignoreBOM: true }));
 			updates.set("docProps/core.xml", updateCoreProperties(corePropsXml, { updateModifiedDate: true }));
 		}
 		if ([...updates.keys()].some((path) => {

package/dist/docx/serializer/runSerializer.js

@@ -9,6 +9,7 @@ import { canReplayEditableImageRawXml } from "../imageRawXml.js";
 import { serializeNonVisualDrawingNames } from "../nonVisualDrawingProps.js";
 import { serializeBoundaryJoins } from "../reviewResolutionProvenance.js";
 import { runHoldsPayload } from "../runPayload.js";
+import { serializeEmptyRunProperties, serializePreviousRunProperties } from "../runPropertyPresence.js";
 import { replayableShapeAlternateContent } from "../shapeAlternateContent.js";
 import { requiredWrapPolygon, serializeWrapPolygon } from "../wrapPolygon.js";
 import { serializeParagraph } from "./paragraphSerializer.js";
@@ -52,7 +53,7 @@ function getUniqueId(id) {
 }
 function serializeRunPropertyChange(change) {
 	const previousRPrXml = serializeTextFormatting(change.previousFormatting) || "<w:rPr/>";
-	return `<w:rPrChange ${serializeTrackedChangeAttributes(change.info)}${serializeBoundaryJoins(change.boundaryJoins)}>${previousRPrXml}</w:rPrChange>`;
+	return `<w:rPrChange ${serializeTrackedChangeAttributes(change.info)}${serializeBoundaryJoins(change.boundaryJoins)}${serializePreviousRunProperties(change.previousFormatting)}>${previousRPrXml}</w:rPrChange>`;
 }
 function serializeRunProperties(formatting, propertyChanges) {
 	const propertyChange = getSingularRunPropertyChange(propertyChanges);
@@ -567,7 +568,7 @@ function serializeRun(run) {
 		const contentXml = serializeRunContent(content);
 		if (contentXml) parts.push(contentXml);
 	}
-	const attrs = serializePreservedAttributes([], run.preservedAttributes);
+	const attrs = serializePreservedAttributes(serializeEmptyRunProperties(run), run.preservedAttributes);
 	return `<w:r${attrs.length > 0 ? ` ${attrs.join(" ")}` : ""}>${parts.join("")}</w:r>`;
 }
 /**

package/dist/docx/serializer/tableSerializer.js

@@ -5,6 +5,7 @@ import { withPreservedChildren } from "../containerChildren.js";
 import { TRANSITIONAL_NAME_BY_STRICT_NAME } from "../strictNames.gen.js";
 import { TABLE_LOOK_FLAGS } from "../tableLook.js";
 import { parseTableCellProperties, parseTableGrid, parseTableProperties, parseTablePropertyExceptions, parseTableRowProperties } from "../tableParser.js";
+import { serializeEmptyRowProperties } from "../tableRowPropertyPresence.js";
 import { OOXML_NAMESPACE_SCOPE, parseXml } from "../xmlParser.js";
 import { serializeBlockCustomXml } from "./blockCustomXmlSerializer.js";
 import { serializeBlockSdt } from "./blockSdtSerializer.js";
@@ -555,7 +556,7 @@ function serializeTableRow(row, serializeParagraph) {
 		carriers: carriersFor(cell),
 		xml: serializeTableCell(cell, serializeParagraph)
 	})), positionedChildren(row.preserved, row.bookmarks), preservedChild), 0));
-	const attrs = serializePreservedAttributes([], row.preservedAttributes);
+	const attrs = serializePreservedAttributes(serializeEmptyRowProperties(row), row.preservedAttributes);
 	return `<w:tr${attrs.length > 0 ? ` ${attrs.join(" ")}` : ""}>${parts.join("")}</w:tr>`;
 }
 /**

package/dist/docx/server/applyDocxXmlPatchProposal.js

@@ -1,3 +1,4 @@
+import { encodeUtf8 } from "../../utils/textEncoding.js";
 import { loadDocxArchive } from "./boundedArchive.js";
 import { FOLIO_DOCX_XML_PATCH_PROPOSAL_PROFILE, InvalidFolioDocxXmlPatchProposalError, evaluateDocxXmlPatchProposal, parseFolioDocxXmlPatchProposal } from "./evaluateDocxXmlPatchProposal.js";
 import { FOLIO_DOCX_CONFORMANCE_PROFILE, validateDocxConformance } from "./validateDocxConformance.js";
@@ -96,9 +97,8 @@ const applyReplacements = async ({ bytes, proposal }) => {
 		writeZipPart({
 			zip,
 			path: replacement.path,
-			data: new TextEncoder().encode(replacement.replacementXml),
+			data: encodeUtf8(replacement.replacementXml),
 			options: {
-				binary: true,
 				date: current.date,
 				comment: current.comment,
 				unixPermissions: current.unixPermissions,

package/dist/docx/server/boundedArchive.js

@@ -1,3 +1,4 @@
+import { decodeText } from "../../utils/textEncoding.js";
 import { compressionRatioLimitFor, countCentralDirectoryRecords, createInflationBudget, exceedsCompressionRatio, getZipEntrySizes, inflateEntryWithinLimits } from "../archiveInflation.js";
 import { FOLIO_XML_RESOURCE_LIMITS, assertXmlResourceLimits, createXmlPackageBudget } from "../xmlResourceLimits.js";
 import { TaggedError } from "better-result";
@@ -170,7 +171,7 @@ const loadDocxArchive = async (bytes, options = {}) => {
 		async readEntryString(path) {
 			const content = await readEntry(path);
 			if (content === null) return null;
-			const xml = new TextDecoder("utf-8", { ignoreBOM: true }).decode(content);
+			const xml = decodeText(content, { ignoreBOM: true });
 			const lower = path.toLowerCase();
 			if (!lower.endsWith(".xml") && !lower.endsWith(".rels")) return xml;
 			const charged = countedParts.has(path);

package/dist/docx/server/evaluateDocxXmlPatchProposal.js

@@ -1,3 +1,4 @@
+import { encodeUtf8 } from "../../utils/textEncoding.js";
 import { getDocxXmlSafetyIssue } from "../xmlSafety.js";
 import { DocxArchiveError } from "./boundedArchive.js";
 import { FolioDocxPackageInspectionError, inspectDocxPackage } from "./inspectDocxPackage.js";
@@ -264,7 +265,7 @@ const evaluateDocxXmlPatchProposal = async ({ bytes, proposal: rawProposal, allo
 			});
 		}
 		if (exceedsPartLowerBound || totalLimitReported) continue;
-		const replacementBytes = new TextEncoder().encode(replacement.replacementXml);
+		const replacementBytes = encodeUtf8(replacement.replacementXml);
 		const exceedsEncodedPart = replacementBytes.byteLength > limits.maxPartBytes;
 		const nextEncodedTotalBytes = encodedTotalBytes + replacementBytes.byteLength;
 		if (exceedsEncodedPart) issues.push({

package/dist/docx/server/inspectDocxPackage.js

@@ -1,3 +1,4 @@
+import { decodeText } from "../../utils/textEncoding.js";
 import { getAttributeAnyPrefix, getLocalName, parseXmlDocument } from "../xmlParser.js";
 import { DocxArchiveError, loadDocxArchive } from "./boundedArchive.js";
 import { Result, TaggedError } from "better-result";
@@ -33,7 +34,10 @@ const decodeXml = (bytes, part) => {
 	if (bytes.length >= 2 && (bytes[0] === 255 && bytes[1] === 254 || bytes[0] === 60 && bytes[1] === 0)) encoding = "utf-16le";
 	else if (bytes.length >= 2 && (bytes[0] === 254 && bytes[1] === 255 || bytes[0] === 0 && bytes[1] === 60)) encoding = "utf-16be";
 	const decoded = Result.try({
-		try: () => new TextDecoder(encoding, { fatal: true }).decode(bytes),
+		try: () => decodeText(bytes, {
+			encoding,
+			fatal: true
+		}),
 		catch: (cause) => new FolioDocxPackageInspectionError({
 			message: `Failed to decode XML package part "${part}"`,
 			code: "xml-decode-failed",

package/dist/docx/server/validateDocxConformance.js

@@ -1,3 +1,4 @@
+import { decodeText } from "../../utils/textEncoding.js";
 import { detectDocxConformanceClass } from "../conformance.js";
 import { DOCX_CONTAINER_TYPES, detectDocxContainerType } from "../encryption/containerFormat.js";
 import { DocxModelValidationError, validateFolioDocumentModel } from "../modelValidation.js";
@@ -131,7 +132,7 @@ const readPartForInspection = async (archive, part) => {
 	} catch (error) {
 		if (!(error instanceof XmlResourceLimitError) || error.limit !== "syntax") throw error;
 		const bytes = await archive.readEntryUint8(part);
-		return bytes === null ? null : new TextDecoder("utf-8", { ignoreBOM: true }).decode(bytes);
+		return bytes === null ? null : decodeText(bytes, { ignoreBOM: true });
 	}
 };
 const validateXmlParts = async (archive, report) => {

package/dist/docx/tableParser.js

@@ -11,10 +11,11 @@ import { captureSdtSiblingMarkers, parseSdtProperties } from "./sdtProperties.js
 import { parseShading } from "./shadingParser.js";
 import { findChildAnySpelling } from "./strictNames.js";
 import { TABLE_LOOK_FLAGS } from "./tableLook.js";
+import { isEmptyRowPropertiesAttribute, retainsEmptyRowProperties } from "./tableRowPropertyPresence.js";
 import { parsePropertyChangeInfo, parseTrackedChangeInfo } from "./trackedChangeInfo.js";
 import { percentageSpelling, transitionalSlotEncoding } from "./transitionalSpelling.js";
 import { captureVerbatimXml } from "./verbatimCapture.js";
-import { WORDPROCESSINGML_NAMESPACE_URIS, cloneElement, findChild, findChildren, findWordprocessingChild, getAttribute, getAttributeByNamespaceUri, getLocalName, mergeXmlnsDeclarations, parseBooleanElement, parseNumericAttribute, parseOnOffAttribute, parseTableMeasurementValue, selectAlternateContentBranch } from "./xmlParser.js";
+import { WORDPROCESSINGML_NAMESPACE_URIS, cloneElement, findChild, findChildren, findWordprocessingChild, getAttribute, getAttributeByNamespaceUri, getChildElements, getLocalName, mergeXmlnsDeclarations, parseBooleanElement, parseNumericAttribute, parseOnOffAttribute, parseTableMeasurementValue, selectAlternateContentBranch } from "./xmlParser.js";
 import { PARSE_WARNING_CODES } from "@stll/docx-core/model";
 //#region src/docx/tableParser.ts
 /**
@@ -637,6 +638,8 @@ function parseTableRowProperties(trPrElement) {
 		context: formatting
 	});
 	if (preserved) formatting.preserved = preserved;
+	const children = getChildElements(trPrElement);
+	if (Object.keys(formatting).length === 0 && children.length > 0 && Object.keys(trPrElement.attributes ?? {}).every((name) => name === "xmlns" || name.startsWith("xmlns:")) && children.every((child) => WORDPROCESSINGML_NAMESPACE_URIS.has(child.namespaceUri ?? "") && REVISION_PROPERTY_CHILDREN.has(getLocalName(child.name)))) return void 0;
 	return withSourceXml(formatting, trPrElement);
 }
 /**
@@ -1099,7 +1102,7 @@ function parseTableRow(trElement, styles, theme, numbering, rels, media, options
 	const exceptionChanges = parseTablePropertyExceptionChanges(tblPrExElement, exceptions);
 	if (exceptionChanges !== void 0) row.tablePropertyExceptionChanges = exceptionChanges;
 	const trPrElement = findChild(trElement, "w", "trPr");
-	const formatting = parseTableRowProperties(trPrElement);
+	const formatting = parseTableRowProperties(trPrElement) ?? (trPrElement && retainsEmptyRowProperties(trElement) ? withSourceXml({}, trPrElement) : void 0);
 	if (formatting) row.formatting = formatting;
 	const rowPropChanges = parseTableRowPropertyChanges(trPrElement, formatting);
 	if (rowPropChanges !== void 0) row.propertyChanges = rowPropChanges;
@@ -1126,8 +1129,8 @@ function parseTableRow(trElement, styles, theme, numbering, rels, media, options
 	const remainingAttributes = attributeRemainder({
 		element: trElement,
 		modelled: NO_MODELLED_ATTRIBUTES
-	});
-	if (remainingAttributes) row.preservedAttributes = remainingAttributes;
+	})?.filter((attribute) => !isEmptyRowPropertiesAttribute(attribute));
+	if (remainingAttributes?.length) row.preservedAttributes = remainingAttributes;
 	return row;
 }
 /**

package/dist/docx/tableRowPropertyPresence.d.ts

@@ -0,0 +1,9 @@
+import { document_d_exports } from "../types/document.js";
+import { XmlElement } from "./xmlParser.js";
+//#region src/docx/tableRowPropertyPresence.d.ts
+declare const EMPTY_ROW_PROPERTIES_ATTRIBUTE = "emptyRowProperties";
+declare const retainsEmptyRowProperties: (element: XmlElement) => boolean;
+declare const serializeEmptyRowProperties: (row: document_d_exports.TableRow) => string[];
+declare const isEmptyRowPropertiesAttribute: ({ namespace, name }: Pick<document_d_exports.PreservedAttribute, "namespace" | "name">) => boolean;
+//#endregion
+export { EMPTY_ROW_PROPERTIES_ATTRIBUTE, isEmptyRowPropertiesAttribute, retainsEmptyRowProperties, serializeEmptyRowProperties };

package/dist/docx/tableRowPropertyPresence.js

@@ -0,0 +1,15 @@
+import { OOXML_NAMESPACES } from "./serializer/partNamespaces.js";
+import { getAttributeByNamespaceUri } from "./xmlParser.js";
+import { EMPTY_PROPERTY_SETS } from "@stll/docx-core/ops";
+//#region src/docx/tableRowPropertyPresence.ts
+const EMPTY_ROW_PROPERTIES_ATTRIBUTE = "emptyRowProperties";
+const HISTORY_NAMESPACES = /* @__PURE__ */ new Set([OOXML_NAMESPACES.folio.uri]);
+const retainsEmptyRowProperties = (element) => getAttributeByNamespaceUri(element, HISTORY_NAMESPACES, EMPTY_ROW_PROPERTIES_ATTRIBUTE) === EMPTY_PROPERTY_SETS.KEEP;
+const serializeEmptyRowProperties = (row) => {
+	if (row.formatting === void 0 || row.structuralChange === void 0 && !row.propertyChanges?.length) return [];
+	const { sourceXml: _source, ...properties } = row.formatting;
+	return Object.keys(properties).length === 0 ? [`folio:${EMPTY_ROW_PROPERTIES_ATTRIBUTE}="${EMPTY_PROPERTY_SETS.KEEP}"`] : [];
+};
+const isEmptyRowPropertiesAttribute = ({ namespace, name }) => namespace === OOXML_NAMESPACES.folio.uri && name === "emptyRowProperties";
+//#endregion
+export { EMPTY_ROW_PROPERTIES_ATTRIBUTE, isEmptyRowPropertiesAttribute, retainsEmptyRowProperties, serializeEmptyRowProperties };

package/dist/docx/unzip.js

@@ -1,4 +1,5 @@
 import { bytesToDataUrl } from "../utils/base64.js";
+import { encodeUtf8 } from "../utils/textEncoding.js";
 import { compressionRatioLimitFor, countCentralDirectoryRecords, createInflationBudget, exceedsCompressionRatio, getZipEntrySizes, inflateEntryWithinLimits, isStoredZipEntry } from "./archiveInflation.js";
 import { DOCX_CONTAINER_TYPES, detectDocxContainerType } from "./encryption/containerFormat.js";
 import { openDocxBuffer } from "./encryption/openEncryptedDocx.js";
@@ -286,7 +287,7 @@ const replaceRawDocxXmlParts = async (content, parts) => {
 		writeZipPart({
 			zip: content.originalZip,
 			path,
-			data: xmlContent
+			data: encodeUtf8(xmlContent)
 		});
 		identitiesChanged = true;
 	}

package/dist/docx/xmlEncoding.js

@@ -1,12 +1,13 @@
+import { decodeText } from "../utils/textEncoding.js";
 //#region src/docx/xmlEncoding.ts
 /** Decode the UTF-8 and UTF-16 encodings XML processors must recognize. */
 const decodeXmlBytes = (bytes) => {
-	if (bytes[0] === 255 && bytes[1] === 254) return new TextDecoder("utf-16le").decode(bytes.subarray(2));
-	if (bytes[0] === 254 && bytes[1] === 255) return new TextDecoder("utf-16be").decode(bytes.subarray(2));
-	if (bytes[0] === 60 && bytes[1] === 0 && bytes[2] === 63 && bytes[3] === 0) return new TextDecoder("utf-16le").decode(bytes);
-	if (bytes[0] === 0 && bytes[1] === 60 && bytes[2] === 0 && bytes[3] === 63) return new TextDecoder("utf-16be").decode(bytes);
+	if (bytes[0] === 255 && bytes[1] === 254) return decodeText(bytes.subarray(2), { encoding: "utf-16le" });
+	if (bytes[0] === 254 && bytes[1] === 255) return decodeText(bytes.subarray(2), { encoding: "utf-16be" });
+	if (bytes[0] === 60 && bytes[1] === 0 && bytes[2] === 63 && bytes[3] === 0) return decodeText(bytes, { encoding: "utf-16le" });
+	if (bytes[0] === 0 && bytes[1] === 60 && bytes[2] === 0 && bytes[3] === 63) return decodeText(bytes, { encoding: "utf-16be" });
 	const utf8Offset = bytes[0] === 239 && bytes[1] === 187 && bytes[2] === 191 ? 3 : 0;
-	return new TextDecoder("utf-8").decode(bytes.subarray(utf8Offset));
+	return decodeText(bytes.subarray(utf8Offset), { encoding: "utf-8" });
 };
 //#endregion
 export { decodeXmlBytes };

package/dist/fonts/fontsourceFaces.js

@@ -1,6 +1,7 @@
 import { bytesToDataUrl } from "../utils/base64.js";
 import { FONT_MAPPING } from "../utils/fontLoader.js";
 import { parseFontFamilyList, resolveFontFamily } from "../utils/fontResolver.js";
+import { decodeText } from "../utils/textEncoding.js";
 //#region src/fonts/fontsourceFaces.ts
 /**
 * The subsets served, in resolution priority: a code point two subsets cover
@@ -78,7 +79,7 @@ const parseRangeToken = (token) => {
 const parseRangeList = (declared) => declared.split(",").flatMap(parseRangeToken);
 const parseDeclaredRanges = (bytes) => {
 	const ranges = /* @__PURE__ */ new Map();
-	const parsed = JSON.parse(new TextDecoder().decode(bytes));
+	const parsed = JSON.parse(decodeText(bytes));
 	if (typeof parsed !== "object" || parsed === null) return ranges;
 	for (const [subset, declared] of Object.entries(parsed)) if (typeof declared === "string") ranges.set(subset, parseRangeList(declared));
 	return ranges;

package/dist/fonts/sfnt/parse.js

@@ -1,3 +1,4 @@
+import { decodeText } from "../../utils/textEncoding.js";
 import { SFNT_TABLE, SFNT_VERSION, fitsInView, readLocaOffsets, readSfntDirectory, readSfntVersion } from "./tables.js";
 import { Result, TaggedError } from "better-result";
 //#region src/fonts/sfnt/parse.ts
@@ -344,17 +345,13 @@ const sanitizePostScriptName = (raw) => {
 	return name.slice(0, MAX_POSTSCRIPT_NAME_LENGTH);
 };
 const decodeUtf16Be = (view, offset, length) => {
-	let text = "";
-	for (let index = 0; index + 1 < length; index += 2) text += String.fromCharCode(view.getUint16(offset + index));
-	return text;
+	return decodeText(new Uint8Array(view.buffer, view.byteOffset + offset, length - length % 2), {
+		encoding: "utf-16be",
+		ignoreBOM: true
+	});
 };
 const decodeMacRomanAscii = (view, offset, length) => {
-	let text = "";
-	for (let index = 0; index < length; index++) {
-		const code = view.getUint8(offset + index);
-		text += code < 128 ? String.fromCharCode(code) : "";
-	}
-	return text;
+	return decodeText(new Uint8Array(view.buffer, view.byteOffset + offset, length).filter((byte) => byte < 128), { encoding: "binary" });
 };
 /**
 * name id 6, preferring the Windows UTF-16BE record over the Macintosh one.

package/dist/fonts/sfnt/subset.js

@@ -1,3 +1,4 @@
+import { encodeBinaryString } from "../../utils/textEncoding.js";
 import { SFNT_TABLE, SFNT_VERSION, fitsInView, readLocaOffsets, readSfntDirectory } from "./tables.js";
 import { Result, TaggedError, panic } from "better-result";
 //#region src/fonts/sfnt/subset.ts
@@ -223,7 +224,7 @@ const assembleSfnt = (tables) => {
 	let dataOffset = 12 + directorySize;
 	let headOffset = -1;
 	for (const table of sorted) {
-		for (let index = 0; index < table.tag.length; index++) view.setUint8(entry + index, table.tag.charCodeAt(index));
+		output.set(encodeBinaryString(table.tag), entry);
 		view.setUint32(entry + 4, checksumOf(table.data));
 		view.setUint32(entry + 8, dataOffset);
 		view.setUint32(entry + 12, table.data.byteLength);

package/dist/fonts/sfnt/tables.js

@@ -1,3 +1,4 @@
+import { decodeText } from "../../utils/textEncoding.js";
 import { Result, TaggedError } from "better-result";
 //#region src/fonts/sfnt/tables.ts
 /**
@@ -65,9 +66,7 @@ const fitsInView = (view, offset, length) => Number.isInteger(offset) && Number.
 /** The four-byte tag at `offset`, or `""` when it does not fit. */
 const readTag = (view, offset) => {
 	if (!fitsInView(view, offset, TAG_LENGTH)) return "";
-	let tag = "";
-	for (let index = 0; index < TAG_LENGTH; index++) tag += String.fromCharCode(view.getUint8(offset + index));
-	return tag;
+	return decodeText(new Uint8Array(view.buffer, view.byteOffset + offset, TAG_LENGTH), { encoding: "binary" });
 };
 /** The leading version word, or `undefined` when the buffer is shorter than it. */
 const readSfntVersion = (bytes) => {

package/dist/fonts/sfnt/woff.js

@@ -1,3 +1,4 @@
+import { encodeBinaryString } from "../../utils/textEncoding.js";
 import { SFNT_VERSION, readSfntVersion, readTag } from "./tables.js";
 import { Result, TaggedError } from "better-result";
 import { inflateSync } from "node:zlib";
@@ -76,7 +77,7 @@ const buildSfnt = (flavor, tables) => {
 	let entry = 12;
 	let dataOffset = 12 + directorySize;
 	for (const table of sorted) {
-		for (let index = 0; index < table.tag.length; index++) view.setUint8(entry + index, table.tag.charCodeAt(index));
+		output.set(encodeBinaryString(table.tag), entry);
 		view.setUint32(entry + 4, table.checksum);
 		view.setUint32(entry + 8, dataOffset);
 		view.setUint32(entry + 12, table.data.byteLength);

package/dist/i18n/messages/ar.gen.js

@@ -362,6 +362,7 @@ const messages = { "folio": {
 		"previous": "النتيجة السابقة",
 		"previousShortcut": "النتيجة السابقة (Shift+Enter)",
 		"replaceAll": "استبدال الكل",
+		"replaceAllSummary": "تم استبدال {replacedCount} وتخطي {skippedCount}",
 		"replaceCurrent": "استبدال النتيجة الحالية",
 		"replacePlaceholder": "نص الاستبدال...",
 		"replaceText": "نص الاستبدال",